It is currently Fri Dec 04, 2020 7:09 am



Post new topic Reply to topic
Author Message
ekstra   ara
 Post subject: Cable Tray Bonding/Grounding
PostPosted: Wed Nov 15, 2017 7:43 pm 
Sparks Level

Joined: Tue Apr 03, 2012 1:28 pm
Posts: 53
Location: Louisville KY
Ok - interesting issue. I am currently overseas in Singapore doing an electrical audit for a US-Based company. This company requires US standards to be followed unless local law is more restrictive. While doing my walk-through, I noticed some questionable cable tray grounding/bonding in compared to what I have seen in other countries that follow NEC.

As an example, a power tray carrying large conductors (equivalent US size of 4/0 range to 300 mcm) with a large fault current available (exceeding 45kA), with a tray that runs approximately 75 feet. The conductor from the tray to the utilization equipment and from the power feed cabinet to the tray is only a size equivalent of 4-6 gauge wire. The tray itself has no bonding wire, thus it depends on the tray splice joints to carry any fault current back to source. The only attempt I see besides the small bonding wires are little copper splices on each section of tray - and I mean little - as in 2mm x 6mm that are no more than 4mm thick at best. This setup is on all trays, be they data trays or 415 volt or 66kV. The trays typically have a larger conductor terminated to each end of the tray run, and this conductor is approximately 2/0 equivalent (all wire sizes in Singapore are in MM2, not in AWG).

Does this sound reasonable? I am accustomed to seeing a 2/0 or 4/0 cable bonded to each tray section, where paint has been removed to ensure proper bonding. Here I see bolted tray sections on powder-coated tray, with a little copper splice.

Thoughts/comments are greatly appreciated. Sorry - no NEC book present with me at the moment to delve into requirements per NEC. Just wanted some general feedback at the moment.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Wed Nov 15, 2017 9:59 pm 
Sparks Level

Joined: Tue Apr 03, 2012 1:28 pm
Posts: 53
Location: Louisville KY
Just to add some more clarity, the upstream c/b inst setting on most of these circuits is around 1200 amps. Per 392.60, 250.96, and 250.102(D), a minimum size appears to be 3/0 copper per NEC 250.122. Best practice on trays is to bond each section of tray with a listed bond/ground clamp, and ensure this bond continues from source to termination point. Table 392.60(A) seems to say that cable trays are NOT allowed to be used as an EGC when OCPD upstream is over 600 amps, thus the tray would require a bonding conductor on its entire length bonded to each section as described above.

Am I on track here, or out in left field?


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Thu Nov 16, 2017 4:44 pm 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
You didn't specify the type of grounding. That makes a major difference if the system grounding is not solidly grounded or low resistance grounding. In an ungrounded or high resistance grounded system, the amount of fault current is very small so heavy grounding cables just shy of that required for continuous currents is not necessary. For instance a 7200 VAC, 10 MVA system with a 25 ampere high resistance ground really only needs #8 grounding and bonding and then only because that's the Code minimum. In such a system it would be pointless to run say 4/0 or 350 MCM grounding conductors when the ground fault current by design will not exceed 25 A. That can be carried for fault purposes by a #14 wire.

The most common but not the only grounding system in the U.S. is solidly grounded. In other countries that is not necessarily the norm and there is quite a bit of variation as to what is the most common or what is required.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Thu Nov 16, 2017 6:32 pm 
Sparks Level

Joined: Tue Apr 03, 2012 1:28 pm
Posts: 53
Location: Louisville KY
Excellent point, Paul. The rooms (multiple locations on same site - large data center) has a 22kV system with resistive grounding with a maximum of 100 amps per the local site EE. The 415V systems has solid grounding with over 45kA SCC available. All the systems are bonded as described and I cannot find any product listing stating the tray is rated as an EGC or bonding conductor.

As well, many of the load circuits - large motors, feeds to server rooms, etc. - are 415V solid-grounded as well.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Sat Nov 18, 2017 7:35 am 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
In those systems I would feel very nervous about using the cable tray as an EGC. I don't particularly like using conduit and cable tray in the first place as an EGC for a very simple reason. Ignoring the whole magnetic conduit/impedance vs. simple resistive argument for moment, consider a 100 A breaker with say 400 V (close to your values). In that case in the event of a fault to get the breaker to trip EVER, we need 400 / 100 = 4 ohms or less of impedance. That is not realistic for most cable tray/conduit runs that I've ever bothered to measure. I'm a huge fan of running an independent grounding conductor for that reason alone.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Sat Nov 18, 2017 1:43 pm 
Sparks Level

Joined: Tue Apr 03, 2012 1:28 pm
Posts: 53
Location: Louisville KY
My thoughts exactly! The on-site EE wasn’t understanding my concern - probably because he had never seen anything different, and from all I can research, it meets Singapore law. Much of their electrical standard is based on British refs, based on when they were occupied by the Brits years back. I was able to find a current British reg on cable trays that requires each cable to have either two layers of insulation or a cable shield, and these trays only require bonding - no mention of fault current, because they “do not expect the tray to become energized.”

I always have and will continue to require a large bare copper conductor bonded to each tray section, as well as to the feeder and load equipment. To me, anything less is just asking for trouble.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Sun Nov 19, 2017 11:34 pm 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
Let me start by stating the obvious: grounding is one of the least understood and more challenging aspects of electrical work. When working on the phase conductors, at least you can take current and voltage readings and generally tell right away when there are problems. Grounding usually carries little or no current at all and just sits there passively doing nothing until the ungrounded part of the system develops a fault at which point it is suddenly and miraculously expected to work as a conductor. So it's no wonder that it often gets taken for granted and that few people actually give it much of any consideration at all except to comply with Codes, so the response from your client isn't all that surprising. Let me give you a little insight though on just how bad it really is.

I just got through doing an equipment grounding survey for a power plant last month. I don't have permission to publish details so I'm just going to say that the plant is not far from the metro DC area. I didn't count but we probably checked at least several hundred grounding conductors plus associated cabling, connections to structures, and so forth. Typically we were seeing ground resistances around 0.1-0.3 ohms or less and grounding was generally excellent everywhere. Even common problem areas like fences and cooling towers were generally trouble free. There were plenty of cable tray installations with a heavy ground cable running right down the tray, terminated every so often to each individual tray or branched off with the phase conductors. But some of the cable tray had the same grade of equipment bonding that you were seeing in Singapore, that had been this way for years..in other words almost nothing at all. And I wasn't even the first inspector that had looked at this before, just the first one to recognize the issue. Mind you the plant E&I superintendent is no dummy and he's going to fix this so the situation is a little different in that regard.

Stepping away from the Codes, actually achieving effective grounding is not as easy as it sounds. There are some surprising pitfalls and the reason that even if everything is done to Code, at least superficially enough to pass inspection, lack of effective grounding is a problem in the U.S. Here are five issues that may surprise you.

The first surprise is that the fact is that long runs of conduit or really any kind of steel or other conductors are simply ineffective as grounding paths. Infinitely long runs are actually forbidden by the performance-based rules of the Code which effectively require proper grounding but determining whether or not conduit as EGC's is effective or not is something that basically nobody does. Quite often especially in large plants they don't really have an effective grounding system but it goes unnoticed as long as everything is properly bonded. The following articles point to the current research and calculations for this purpose. I have some old tables from the original Soares information that included the calculations behind i. I've used this as training materials for technicians for years showing how far they can push things before a ground conductor goes from optional to required. It's really quite easy to figure out once you're aware of the issue.

https://iaeimagazine.org/magazine/2004/ ... tudy-gemi/

https://steeltubeinstitute.org/steel-co ... unding.pdf

The second surprise is that assuming that the grounding connections are properly made, problems develop surprisingly quickly. Many electricians fail to use a hammer and screwdriver to properly penetrate the paint and bury conduit luck nut teeth in the metal surface they are intended to engage. Sometimes these connections pass at least for a time but corrosion quickly clears them out. And don't get me started with flexible metallic conduits...that stuff is good for about 6 months to a year before it is utterly useless as an EGC. I know because in the mining industry we have to test EGC's annually in some operations, and flexible conduits fail consistently the following year. Rigid conduit lasts a lot longer before it doesn't make good contact anymore from corrosion. Worse still since we are mixing different metals and adding the element of electrical currents, galvanic corrosion is not only present but it is accelerated by basically the reverse process of galvanic protection. So where problems exist, they fail very quickly. This all goes unnoticed in most plants because there isn't a regulatory requirement to test. But if you test, there are always conduits here and there every year where eventually a supplemental grounding conductor will have to be added. There is an argument that if the current is high enough as in a fault it will blow right through corrosion but it should also be remembered that this occurs at a time when the voltage drop is very small and unlikely to "burn a path" as a result. This is very different from lightning protection where poorly made connections are likely to work anyways.

A third surprise that is basically unnoticed because the boxes are legal (though I have no idea how) is knockout rings. Particularly NEMA 1 enclosures but even some NEMA 3R's come with these cursed things. A knockout ring is where you have a prepunched hole for a conduit entry but it has several rings, one inside the other. The manufacturer leaves two very tiny metal strips left between the ring and the box holding the rings in place. From a conductivity point of view though through steel or aluminum, there is no way that two tiny 1/8" or smaller strips of metal are going to conduct fault current, particularly if we consider that the narrow slot causes voltage stress that "pinches off" the current flow even more. The result is that knockout rings should either be not used or knocked out to the outer hole, never an intermediate size. It will pass a simple ohm check but cannot sustain fault currents at all. I've seen damage incurred at knockout rings after a ground fault that basically melted and collapsed the conduit fitting right out of the box despite minimal damage elsewhere.

The fourth surprise is going to sound more like a stroll through the Twilight Zone and if you are trying to deal with finicky ground continuity monitors, it truly is. If everything is bonded and not insulated from the building structure and this is generally the Code requirement, particularly with very large or long structures, our concepts of grounding and bonding tend to fall apart. The resistance through the Earth is inversely proportional to the distance between ground electrodes and the rate that it overcomes the resistance of the intentional equipment grounding conductors is surprisingly short and happens in a few thousand feet at most. This means that the ground path, particularly in multigrounded systems, can pass through Earth via a structural ground at one point, then re-enter at another point in an EGC that may not even be associated with the phase conductors where the ground fault exists, and then pass back to the transformer on this alternate route. This is particularly problematic in mining especially when trying to make regulatory requirements on ground continuity monitors actually work when they are based on faulty logic.

The fifth and final surprising issue has to do with mechanically strong connections. Bonding generally isn't done to meet any kind of electrical force requirements and it is often not given much attention at all. The forces developed during a fault are enormous. Recognizing this has generally changed the construction of temporary equipment grounds in the utility business for instance. This video should give you some practical idea of the problem and why the generally loose and mechanically poor connections that are done for EGC's are never going to hold up under fault conditions:
https://www.youtube.com/watch?v=hG2p9N1MHlk&t=369s

So the situation you describe really doesn't surprise me at all. And I'm not even going to throw them under the bus when we have all kinds of blatant problems with effective grounding right here in North America where we should know better, and some of the issues that are out there are recognized at least by a minority of people in the industry, but have gone unnoticed and are not addressed at all in the Codes except indirectly in some of the performance-based language.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Cable Tray Bonding/Grounding
PostPosted: Mon Nov 20, 2017 7:16 am 
Sparks Level

Joined: Tue Apr 03, 2012 1:28 pm
Posts: 53
Location: Louisville KY
So funny to see this fantastic old video. I use this in all of my Medium Voltage classes in just the context you describe. It is one of the best videos out there on the importance of grounding and equipotential bonding. There is a re-made version, but they slaughtered the video trying to modernize it, in my opinion.

Love your detailed response. I'm a stickler on grounding and bonding, which led me to be so stubborn on the find in Singapore. I can't, unfortunately, force them to fix it, as I am only being paid as an outside party to audit. I would love your permission to forward your comments to the local EE to attempt to broaden his horizons a bit in this area. He is dead-set currently that it meets the law, which unfortunately, it apparently does.

Quick story on the use of conduit as an EGC. When I was a young electrician, fresh out of the Naval Nuclear Power program, my first job was at a large aluminum company. Back in the day, the company's engineering standard use to allow conduit to serve as the EGC. That policy has since been changed, but being 60 to 70 years old, the site had many conduits with no internal EGC. I had been at the site only 2 months, and one day, while standing in a large crane bay, began hearing a popping noise above in the 45 foot overhead. I glanced up just in time to see a 2" conduit coupling glow red, and subsequently explode, and watched this process re-occur the entire length of the conduit run, which was about 150 feet. The fault destroyed the entire conduit run and cables, and we spent several days ripping out and replacing the 2" run of conduit and cables, this time WITH a ground wire in the conduit. Taught me a great lesson about grounding and bonding, and about EXCEEDING the NEC when it comes to system functionality.

Your time and detailed reply are very much appreciated!


Top
 Profile Send private message  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 8 posts ] 

All times are UTC - 7 hours


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
© 2019 Arcflash Forum / Brainfiller, Inc. | P.O. Box 12024 | Scottsdale, AZ 85267 USA | 800-874-8883